src/third_party/WebKit/Source/core/html/parser/BackgroundHTMLParser.h

   1 /*
   2  * Copyright (C) 2013 Google, Inc. All Rights Reserved.
   3  *
   4  * Redistribution and use in source and binary forms, with or without
   5  * modification, are permitted provided that the following conditions
   6  * are met:
   7  * 1. Redistributions of source code must retain the above copyright
   8  *    notice, this list of conditions and the following disclaimer.
   9  * 2. Redistributions in binary form must reproduce the above copyright
  10  *    notice, this list of conditions and the following disclaimer in the
  11  *    documentation and/or other materials provided with the distribution.
  12  *
  13  * THIS SOFTWARE IS PROVIDED BY GOOGLE INC. ``AS IS'' AND ANY
  14  * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  15  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
  16  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL GOOGLE INC. OR
  17  * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
  18  * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
  19  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
  20  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
  21  * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
  22  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
  23  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  24  */
  25
  26 #ifndef BackgroundHTMLParser_h
  27 #define BackgroundHTMLParser_h
  28
  29 #include "core/html/parser/BackgroundHTMLInputStream.h"
  30 #include "core/html/parser/CompactHTMLToken.h"
  31 #include "core/html/parser/HTMLParserOptions.h"
  32 #include "core/html/parser/HTMLPreloadScanner.h"
  33 #include "core/html/parser/HTMLSourceTracker.h"
  34 #include "core/html/parser/HTMLTreeBuilderSimulator.h"
  35 #include "core/html/parser/XSSAuditorDelegate.h"
  36 #include "wtf/PassOwnPtr.h"
  37 #include "wtf/WeakPtr.h"
  38
  39 namespace WebCore {
  40
  41 class HTMLDocumentParser;
  42 class SharedBuffer;
  43 class XSSAuditor;
  44
  45 class BackgroundHTMLParser {
  46     WTF_MAKE_FAST_ALLOCATED;
  47 public:
  48     struct Configuration {
  49         HTMLParserOptions options;
  50         WeakPtr<HTMLDocumentParser> parser;
  51         OwnPtr<XSSAuditor> xssAuditor;
  52         OwnPtr<TokenPreloadScanner> preloadScanner;
  53         OwnPtr<TextResourceDecoder> decoder;
  54     };
  55
  56     static void start(PassRefPtr<WeakReference<BackgroundHTMLParser> >, PassOwnPtr<Configuration>);
  57
  58     struct Checkpoint {
  59         WeakPtr<HTMLDocumentParser> parser;
  60         OwnPtr<HTMLToken> token;
  61         OwnPtr<HTMLTokenizer> tokenizer;
  62         HTMLTreeBuilderSimulator::State treeBuilderState;
  63         HTMLInputCheckpoint inputCheckpoint;
  64         TokenPreloadScannerCheckpoint preloadScannerCheckpoint;
  65         String unparsedInput;
  66     };
  67
  68     void appendRawBytesFromParserThread(const char* data, int dataLength);
  69
  70     void appendRawBytesFromMainThread(PassOwnPtr<Vector<char> >);
  71     void setDecoder(PassOwnPtr<TextResourceDecoder>);
  72     void flush();
  73     void resumeFrom(PassOwnPtr<Checkpoint>);
  74     void startedChunkWithCheckpoint(HTMLInputCheckpoint);
  75     void finish();
  76     void stop();
  77
  78     void forcePlaintextForTextDocument();
  79
  80 private:
  81     BackgroundHTMLParser(PassRefPtr<WeakReference<BackgroundHTMLParser> >, PassOwnPtr<Configuration>);
  82     ~BackgroundHTMLParser();
  83
  84     void appendDecodedBytes(const String&);
  85     void markEndOfFile();
  86     void pumpTokenizer();
  87     void sendTokensToMainThread();
  88     void updateDocument(const String& decodedData);
  89
  90     WeakPtrFactory<BackgroundHTMLParser> m_weakFactory;
  91     BackgroundHTMLInputStream m_input;
  92     HTMLSourceTracker m_sourceTracker;
  93     OwnPtr<HTMLToken> m_token;
  94     OwnPtr<HTMLTokenizer> m_tokenizer;
  95     HTMLTreeBuilderSimulator m_treeBuilderSimulator;
  96     HTMLParserOptions m_options;
  97     WeakPtr<HTMLDocumentParser> m_parser;
  98
  99     OwnPtr<CompactHTMLTokenStream> m_pendingTokens;
 100     PreloadRequestStream m_pendingPreloads;
 101     XSSInfoStream m_pendingXSSInfos;
 102
 103     OwnPtr<XSSAuditor> m_xssAuditor;
 104     OwnPtr<TokenPreloadScanner> m_preloadScanner;
 105     OwnPtr<TextResourceDecoder> m_decoder;
 106     DocumentEncodingData m_lastSeenEncodingData;
 107 };
 108
 109 }
 110
 111 #endif